On 26 December 2015 at 11:49, Eli Zaretskii <[email protected]> wrote: >> Date: Sat, 26 Dec 2015 11:01:15 +0000 >> From: Gavin Smith <[email protected]> >> Cc: Per Bothner <[email protected]>, Texinfo <[email protected]> >> >> On 26 December 2015 at 10:58, Eli Zaretskii <[email protected]> wrote: >> >> If the duplication of the encoding information is a problem, maybe the >> >> Emacs mode for Texinfo could recognize a "@documentencoding" command. >> > >> > But @documentencoding doesn't necessarily mean the file is in that >> > encoding, does it? >> >> I thought it did? That's what the manual seems to say: >> >> "The '@documentencoding' command declares the input document encoding, >> and can also affect the encoding of the output." > > A plain-ASCII document could produce a UTF-8 encoded Info manual, no?
This is subtle. It's true that a plain ASCII document is encoded in UTF-8 (because ASCII is a subset of UTF-8), but the effect of a "@documentencoding UTF-8" directive in such a case wouldn't be on the interpretation of the input, but on the output for Info and HTML. XML and Docbook output always uses UTF-8, and DVI or PDF output doesn't have an encoding. I wouldn't see anything wrong with an option being added to texi2any to specify the output encoding for Info or HTML in spite of any @documentencoding command. It's even possible that the default encoding of the output could change in the future: for example, if @documentencoding is in some strange encoding, and the user is using a UTF-8 locale, it would be likely better for the output encoding to be UTF-8.
